[GitHub] trafficserver issue #807: TS-4672: Add lifecycle events to hook-trace exampl...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/807 I wonder if we should move this out of the examples tree? It seems useful as a debugging tool, and most people probably don't look in examples for useful tools :). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #808: TS-4673: Remove unused HttpSM::historical_action.
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/808 :+1: --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #805: TS-4595: TSRuntimeDirGet
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/805 Unless @yatsukhnenko also wants to fix the thread safety issue on this, and all APIs? If not, file another Jira IMO. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #805: TS-4595: TSRuntimeDirGet
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/805 Sounds good, can you file a new Jira for these issues? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #809: TS-4674: Remove useless assert statement
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/809 +1, although I'm trusting your call here on the internals and changes already made :). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #804: TS-4624: use the server UUID in the Via header
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/804 @yatsukhnenko What do you think of a future feature (config) to disable this Via checking completely? If I know there's not chance of a loop, why bother checking the Via header? I guess one could just nuke the Via header on the incoming request, but there might be other reasons to keep it? This is completely unrelated to this PR though, just thinking out loud. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #804: TS-4624: use the server UUID in the Via header
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/804 Did a quick test of this, and it looks good: Via: http/1.1 fedora.ogre.com[f6f5162d-5a02-4d10-ac6e-35aede508fe5] (ApacheTrafficServer/7.0.0) --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #804: TS-4624: use the server UUID in the Via header
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/804 I suspect this failed for other reasons, trying again [approve ci]. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #804: TS-4624: use the server UUID in the Via header
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/804 I checked the core on this FreeBSD build, and it seems completely unrelated to pretty much anything :). ``` (gdb) bt #0 0x in ?? () #1 0x006ec34a in EThread::process_event (this=0x804107000, e=0x803f2c880, calling_code=) at I_Continuation.h:153 #2 0x006ec8ec in EThread::execute (this=0x804107000) at ../../../iocore/eventsystem/UnixEThread.cc:245 #3 0x006ebb82 in spawn_thread_internal (a=0x803ef5700) at ../../../iocore/eventsystem/Thread.cc:84 #4 0x0008027514f5 in pthread_create () from /lib/libthr.so.3 #5 0x in ?? () ``` --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #804: TS-4624: use the server UUID in the Via header
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/804 Yeah, file a Jira probably. I could have sworn we've had issues before where loop detection kicked in when it shouldn't (e.g. you legitimately proxy to yourself). But I can find an issue on it, maybe you can take a look too. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #804: TS-4624: use the server UUID in the Via hea...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/804 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #805: TS-4595: TSRuntimeDirGet
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/805 @yatsukhnenko As James suggested, can you just add a line or two to doc/developer-guide/api/functions/TSInstallDirGet.en.rst instead of creating a new file? This would be my preference as well. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #814: TS-4689 Assert if Machine UUID init fails
GitHub user zwoop opened a pull request: https://github.com/apache/trafficserver/pull/814 TS-4689 Assert if Machine UUID init fails You can merge this pull request into a Git repository by running: $ git pull https://github.com/zwoop/trafficserver TS-4689 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/trafficserver/pull/814.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #814 commit dbc39e0576d77b8ca4f0453b7c7448f3958ca813 Author: Leif Hedstrom <zw...@apache.org> Date: 2016-07-19T16:38:25Z TS-4689 Assert if Machine UUID init fails --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #814: TS-4689 Assert if Machine UUID init fails
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/814 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #813: TS-4652 Fixes log field initialization, pre...
GitHub user zwoop opened a pull request: https://github.com/apache/trafficserver/pull/813 TS-4652 Fixes log field initialization, previously crippled You can merge this pull request into a Git repository by running: $ git pull https://github.com/zwoop/trafficserver TS-4652 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/trafficserver/pull/813.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #813 commit b9d9130349b155198c1c3c35adc46db565e32392 Author: Leif Hedstrom <zw...@apache.org> Date: 2016-07-19T16:17:54Z TS-4652 Fixes log field initialization, previously crippled --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #810: TS-4671: Allow multicert line with action b...
Github user zwoop commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/810#discussion_r71376108
--- Diff: iocore/net/SSLUtils.cc ---
@@ -2010,8 +2013,9 @@ ssl_extract_certificate(const matcher_line
*line_info, ssl_user_config
}
}
if (!sslMultCertSettings.cert) {
-Warning("missing %s tag", SSL_CERT_TAG);
-return false;
+// TS-4679: It is ok to be missing the cert. At least if the action
is set to tunnel
--- End diff --
I think you should just nuke this commented code, and maybe add a comment
above about the special lenience case for the tunneling (where you check if
sslMultCertSettings.cert is NULL or not).
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
[GitHub] trafficserver issue #805: TS-4595: TSRuntimeDirGet
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/805 Great, thanks! [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #805: TS-4595: TSRuntimeDirGet
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/805 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #813: TS-4652 Fixes log field initialization, previously...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/813 Let me take a look, but I was going for the simplest solution that works, with minimal risk. It's pretty finicky on the order :) --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #815: TS-4680: thread safe initialization in TS*DirGet()...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/815 Interesting, so e.g. static char *runtimedir = RecConfigReadRuntimeDir(); would be thread safe. That seems like a no-brainer to me :). +1 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #813: TS-4652 Fixes log field initialization, previously...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/813 Ah, yeah, I'm pretty sure that entire if() can be safely eliminated (famous last words ...). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #813: TS-4652 Fixes log field initialization, previously...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/813 @jpeach @bryancall check the update please. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #815: TS-4680: thread safe initialization in TS*DirGet()...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/815 If I understand C++ at all, I don't think that example is thread safe? But, moving it outside (file scope) would be. No? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #815: TS-4680: thread safe initialization in TS*DirGet()...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/815 Gotcha. Well, you showed me C++11 has this requirement for thread safety, and we've said that v7.0.0 will require C++11 support, so +1 on that. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #815: TS-4680: thread safe initialization in TS*DirGet()...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/815 @jpeach Wdyt? Would TLS make more sense here? It would duplicate those static strings, but avoids locks, no ? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #817: TS-4667 Uses the WKS in the gzip plugin
GitHub user zwoop opened a pull request: https://github.com/apache/trafficserver/pull/817 TS-4667 Uses the WKS in the gzip plugin Instead of just using strings, like "deflate". Albeit, it's functionally the same, this is a bad pattern that we should discourage (using the WKS's consistently is better performance, and better to read). You can merge this pull request into a Git repository by running: $ git pull https://github.com/zwoop/trafficserver TS-4667 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/trafficserver/pull/817.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #817 commit 82a7db79306949a85cc7fe9f7bde48b849e2860a Author: Leif Hedstrom <zw...@apache.org> Date: 2016-07-20T15:37:37Z TS-4667 Uses the WKS in the gzip plugin --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #818: TS-4683 Adds better error handling on confi...
GitHub user zwoop opened a pull request: https://github.com/apache/trafficserver/pull/818 TS-4683 Adds better error handling on config problems You can merge this pull request into a Git repository by running: $ git pull https://github.com/zwoop/trafficserver TS-4683 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/trafficserver/pull/818.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #818 commit 983e4c469805950d9c8a68bc1de35aabf112ed7e Author: Leif Hedstrom <zw...@apache.org> Date: 2016-07-19T03:45:21Z TS-4683 Adds better error handling on config problems --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #644: [TS 4437] Add new limit rate example plugin
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/644 Aww, don't close it. I don't think anything absolutely wrong here, we should land this. There are things that needs to be fixed, but they are pretty minor (like, squashing, clang-format, typos like "avial" which likely should be "avail" etc. I think you should reopen this. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #792: TS-4650: cachekey: not thread safe
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/792 [approve ci] @gtenev Please review. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #803: TS-4584: Resource leak, CID 1254818 and 1254809
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/803 This fails on clang-format, please re-run and push --force. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #803: TS-4584: Resource leak, CID 1254818 and 1254809
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/803 add to whitelist --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #804: TS-4624: use the server UUID in the Via header
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/804 It's an inefficiency for starters (you go through a C-> C++ wrapper layer, for no good reason (remember, the public APIs are all C, whereas in the core, you can use all the C++ objects and methods directly). In this case, it's marginal, but there are other places where it would be inefficient. And, it would add dependencies to the C-APIs in the C++ core which are undesirable in some cases. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #644: [TS 4437] Add new limit rate example plugin
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/644 Hmmm, well, if it doesn't work, then it's a moot point. I know people have implemented rate limiting in plugins, it was just never open sourced. If you don't feel strongly about it, closing it is fine. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #840: TS-4711: Remove proxy.config.log.custom_logs_enabl...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/840 :+1: --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #841: TS-4720 correctly check if requests have a body
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/841 Wonder if there's some easy way to add this check to the synthetic server (i.e. the regressions as run with traffic_server -R 1)? Also, this might be a back port candidate for 6.2.x (our current LTS). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #845: TS-3245: Allow multiple plugins to safely use geto...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/845 For @PSUdaemon . --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #836: TS-4710 make srv_enabled transaction overrideable
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/836 Weird: REGRESSION TEST SDK_API_OVERRIDABLE_CONFIGS started FATAL: InkAPI.cc:8354: failed assertion `sdk_sanity_check_null_ptr((void *)name) == TS_SUCCESS` --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #837: TS-4706 Truncated SNI name during escalation
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/837 :+1: --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #834: TS-4707 : Parent Consistent Hash Selection - add f...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/834 @jrushford Please review. [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #833: TS-3474: HTTP/2 Server Push support
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/833 Trying another build, for some reason builds are not stable now [approve ci]. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #823: TS-4697: free MIOBuffer if fails on ipallow check.
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/823 Certainly seems reasonable to me. [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #822: TS-4698: Adds an API call to identify WebSocket re...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/822 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #701: TS-4522: check EPIPE instead r==0
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/701 @oknet How do you want to proceed with this? Are you making an update to the PR to address some of the review suggestions? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #771: TS-4612: Proposal: InactivityCop Optimize
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/771 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #823: TS-4697: free MIOBuffer if fails on ipallow check.
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/823 @oknet This failed clang-format, please fix and push update :). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #822: TS-4698: Adds an API call to identify WebSocket re...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/822 Cool. Couple of things: 1) Please email dev@ with a short description for the API proposal. For some details on the process of adding new APIs, see https://cwiki.apache.org/confluence/display/TS/API+Review+Process. 2) Can you add some documentation as well? It doesn't have to be particularly involved, but perhaps add it to TSHttpTxnIsInternal.en.rst? We prefer to group related APIs together in a single man page type file. 3) I don't think we have any existing test harnesses or regression tests for web sockets, but something to consider later. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #766: TS-4614: avoid e->schedule_in for dummy event
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/766 @shinrich Should we land this? Also, is this a back port candidate? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #753: Proposal: NetVC Context
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/753 @shinrich Should we continue with this PR? @oknet there are now merge conflicts, can you please rebase (and remember to re-run clang-format too :). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #802: TS-4583: Null-pointer dereference after check
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/802 @alhonen Thoughts on the concerns from James? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #769: TS 4593: Extend ip_allow.config to filter destinat...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/769 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #819: TS-4311 Removes support for SPDY completely
Github user zwoop commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/819#discussion_r71749709 --- Diff: lib/ts/ink_config.h.in --- @@ -127,5 +127,6 @@ #define TS_BUILD_CANONICAL_HOST "@host@" #define TS_BUILD_DEFAULT_LOOPBACK_IFACE "@default_loopback_iface@" +/* clang-format off */ --- End diff -- Thanks, yeah, should be on (copy-paste failure). So yeah, we don't necessarily need this, however, I ended up screwing this up myself since I edited this file, and wanted to make sure clang-format didn't change anything. And then of course, the entire file gets foobared. I think it's a good measure to prevent zwoop from screwing it up again. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #818: TS-4683 Adds better error handling on config probl...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/818 Made the changes per Gancho's suggestions. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #815: TS-4680: thread safe initialization in TS*D...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/815 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #870: TS-4548: Fix LogHost leak on config error path.
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/870 :+1: --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #859: TS-4724: Modified ts_lua plugin to provide ...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/859 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #868: TS-4719: Make AIO threads interleave memory...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/868 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #862: TS-4691 Removes a bogus assertion which can...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/862 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #858: TS-4184 Makes -std=c++11 mandatory
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/858 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #860: TS-4695 Removes the --enable-cppapi build option
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/860 Sigh, freebsd regression arbitrarily failed it seems ... try again [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #867: TS-2048 Avoids scheduling CFLUS compressor ...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/867 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #861: TS-4360 Renames NPN_ defines to ALPN_
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/861 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #863: TS-4750: Fix Connection Leak warnings.
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/863 @shinrich Lets do it! :) --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #869: TS-4716: Install manpages in $PREFIX/share/man.
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/869 I'm +1 on this, would be curious if the original Jira poster can / have verify ? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #866: TS-2237: Fix double encoding of URLs in squid logs...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/866 @maskit Are you saying the patch as is is not good / sufficient? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #644: [TS 4437] Add new limit rate example plugin
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/644 I'm not sure if anyone has had the cycles to help debug this :-/ It might be worth emailing dev@ with a concrete question / request for help. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1397: TS-1768: AC_SEARCH_LIBS vs. AC_CHECK_LIB
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1397 Hmmm, not sure, it is for me: ``` [root@fedora ats]# ldd bin/traffic_top | grep -i hwloc libhwloc.so.5 => /lib64/libhwloc.so.5 (0x7f29395a7000) ``` That's from the install dir. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1418: Consistently use ink_atoi() (or one of the equiva...
GitHub user zwoop opened an issue: https://github.com/apache/trafficserver/issues/1418 Consistently use ink_atoi() (or one of the equivalent functions we provide) (This is moved from https://issues.apache.org/jira/browse/TS-1773). Right now, we have a hotchpotch of various "string to integer" conversions happening in the core. We have a main one, ink_atoi() / ink_atoui64(). We should unify the code around this, such that we do not use atoi/strtol and other implementations. @SolidWallOfCode Also suggests that we make these APIs a little more feature rich, including: - Can take a pointer and length, not require null termination of input string - Returns the number of characters parsed, or a pointer to the first unparsed character (like strtol) - Works with 64 bit numbers I think we already have most of this covered with the existing implementations in ParseRules.cc: ```C int64_t ink_atoi64(const char *); uint64_t ink_atoui64(const char *); int64_t ink_atoi64(const char *, int); ink_atoi(const char *str); ink_atoi(const char *str, int len) ink_atoui(const char *str) ``` So, what I think is left doing here is (as the original Jira suggested): 1. Go through the entire core code base, and find usage of string-to-int conversions (such as ::atoi, ::strtol, and possibly other implementations), and change them to use one of the above APIs as appropriate. 1. Optional: @SolidWallOfCode Expresses a wish to refactor the implementations in ParseRules.cc / ParseRules.h, such that there is less code duplication (i.e. refactor with a single shared base function or some such). 1. Optional: Expose two generic, public APIs to ts/ts.h. These could/should be three different pull requests IMO. #2 is where we sort of got lost I think. But @SolidWallOfCode posted a suggested public API, that covers all potential use cases in a single, nice C-API (edited here for correctness on the naming): ```C int64_t TSatoi(const const* src, size_t n, const char** end, int radix); uint64_t TSatou(const const* src, size_t n, const char** end, int radix); ``` The implementation (in InkAPI.cc) for these two APIs would simply be wrappers over the existing ink_atoi() based APIs. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1407: Fix potential spin loop in perl management...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/1407 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1408: Prune some unused library dependencies
Github user zwoop commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/1408#discussion_r99240250 --- Diff: iocore/cache/Cache.cc --- @@ -1041,12 +1041,12 @@ CacheProcessor::cacheInitialized() case CACHE_COMPRESSION_FASTLZ: break; case CACHE_COMPRESSION_LIBZ: -#if !TS_HAS_LIBZ +#ifndef HAVE_ZLIB_H --- End diff -- Agreed. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1404: TS-1403 retry safe methods in case of server fail...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1404 Looks great. Two issues: 1) Run clang-format on it :). 2) On FreeBSD, it fails with ``` ../../plugins/experimental/ts_lua/ts_lua_http_config.c:244:53: error: '__ts_lua_http_config_vars_static_assert' declared as an array with a negative size extern char __ts_lua_http_config_vars_static_assert[NUM_HTTP_CONFIG_VARS == TS_CONFIG_LAST_ENTRY + 1 ? 0 : -1]; ^ ../../plugins/experimental/ts_lua/ts_lua_http_config.c:243:30: note: expanded from macro 'NUM_HTTP_CONFIG_VARS' #define NUM_HTTP_CONFIG_VARS (sizeof(ts_lua_http_config_vars) / sizeof(ts_lua_http_config_vars[0])) ^ 1 error generated. ``` --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1404: TS-1403 retry safe methods in case of server fail...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1404 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1401: Segfault in write_to_net_io with 7.1.x
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/issues/1401 you didn't see this crasher in 7.0.0 right ? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1402: Crash in InactivityCop::check_inactivity with dea...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/issues/1402 If possible, can you test with 7.0 or 7.1x (branch) and see if it happens there too ? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1388: Fixes for building with OpenSSL 1.1
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1388 [approve ci] to get the clang-analyzer build --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1388: Fixes for building with OpenSSL 1.1
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1388 @jablko do we need two commits here? Or can they be squashed? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1381: Lib Scalar
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1381 This still has 1) clang-format issues 2) Compile issues: ``` In file included from ../../../lib/ts/test_Scalar.cc:24: /var/jenkins/workspace/freebsd-github/src/BUILD/../lib/ts/Scalar.h:68:34: error: member reference base type 'S *' (aka 'Scalar<N, C, T> *') is not a structure or union return static_cast(this).operator+=(static_cast(n)); ~~^~~ /var/jenkins/workspace/freebsd-github/src/BUILD/../lib/ts/Scalar.h:74:34: error: member reference base type 'S *' (aka 'Scalar<N, C, T> *') is not a structure or union return static_cast(this).operator-=(static_cast(n)); ~~^~~ /var/jenkins/workspace/freebsd-github/src/BUILD/../lib/ts/Scalar.h:80:34: error: member reference base type 'S *' (aka 'Scalar<N, C, T> *') is not a structure or union return static_cast(this).operator*=(static_cast(n)); ~~^~~ /var/jenkins/workspace/freebsd-github/src/BUILD/../lib/ts/Scalar.h:86:34: error: member reference base type 'S *' (aka 'Scalar<N, C, T> *') is not a structure or union return static_cast(this).operator/=(static_cast(n)); ~~^~~ /var/jenkins/workspace/freebsd-github/src/BUILD/../lib/ts/Scalar.h:912:57: error: default initialization of an object of const type 'const ApacheTrafficServer::detail::tag_label_B' requires a user-provided default constructor static ApacheTrafficServer::detail::tag_label_B const b; ``` Also, probably want to rebase again, it's been a while so the list of changes is long ;). --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1335: Deadlock in HostDB
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/issues/1335 @randall reports that this happens with sync off as well :-/ --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1415: isuue #1399 add back partial code of tcp_congesti...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1415 Oh yeah, if this is not a back port from master, then this should be changed / closed, and replaced with a PR against master. We always commit on master, and then suggest back ports to the RMs if needed. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1415: isuue #1399 add back partial code of tcp_congesti...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1415 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1415: isuue #1399 add back partial code of tcp_congesti...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1415 You can ignore the clang-analyzer build failures, since 6.2.x branch is busted. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1392: TS-2095: TS_HAS_LIBZ and TS_HAS_LZMA are always f...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1392 Should we land this? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1421: Segmentation fault on TLS when destination server...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/issues/1421 @jacksontj Ping on this one? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1421: Segmentation fault on TLS when destination server...
GitHub user zwoop opened an issue: https://github.com/apache/trafficserver/issues/1421 Segmentation fault on TLS when destination server issues Reset Creating this issue, copying from the original Jira to capture all the details. On HEAD as of 16DEC2016 (102b66b9730780289ef71fa1ee33214cbcbcdde2) When in Forward proxy mode, connecting to an HTTPS site, where the destination site responds with TCP RESET, Traffic Server will restart due to Sig 11: Segmentation fault. If the upstream site instead silently drops the response, ATS will successfully proceed with 502 (Tunnel Connection Failed). No proxy chaining involved, direct: client -> ATS -> destination. Test machine: Debian Jessie (8.6) x86_64 To replicate, make an HTTPS request - via ATS in forward-proxy-mode - to a local (non-firewalled) server, with nothing running on TCP/443. Watch manager.log for near immediate "[LocalManager::pollMgmtProcessServer] Server Process terminated due to Sig 11: Segmentation fault" followed by "[Alarms::signalAlarm] Server Process was reset" The issue was introduced in c1ac5f8bf87fd4bc3a8e06507219970d83965acd TS-4796 No promises - I'll see what I can do. I've tracked it down to the change to line 508 in iocore/net/UnixNet.cc https://github.com/apache/trafficserver/commit/c1ac5f8bf87fd4bc3a8e06507219970d83965acd#diff-24ea729d4646e63a76dd276a189da30cR508 This is all from "Craig B" in TS-5098. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1357: TS-5022: reduce memory allocation in clientcert l...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1357 @SolidWallOfCode Ping? Also, seems this needs to be rebased now. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1420: Improve error message for custom log field...
Github user zwoop commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/1420#discussion_r99900589
--- Diff: proxy/logging/LogBuffer.cc ---
@@ -615,10 +618,14 @@ LogBuffer::resolve_custom_entry(LogFieldList
*fieldlist, char *printf_str, char
}
bytes_written += res;
-field = fieldlist->next(field);
+lastField = field;
+field = fieldlist->next(field);
} else {
Note("There are more field markers than fields;"
- " cannot process log entry");
+ " cannot process log entry '%.*s'. Last field = '%s'
printf_str='%s' pos=%d/%d count=%d alt_printf_str='%s'",
+ bytes_written, write_to, lastField == nullptr ? "*" :
lastField->symbol(),
--- End diff --
Hmmm, maybe it's completely unrealistic, but couldn't lastField be NULL
here ?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
[GitHub] trafficserver issue #1422: YTS-614: Avoid crash due to error during POST req...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1422 @SolidWallOfCode Is this something you feel comfortable with, and should be a 7.1.x back port? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1412: Assertion in HttpSM::tunnel_handler_post
GitHub user zwoop opened an issue: https://github.com/apache/trafficserver/issues/1412 Assertion in HttpSM::tunnel_handler_post I think we've seen this before, but filing this again to make sure I don't lose the info: ``` (gdb) bt #0 0x74fed1d7 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56 #1 0x74fee8c8 in __GI_abort () at abort.c:90 #2 0x77bb89c8 in ink_abort (message_format=0x77bca590 "%s:%d: failed assertion `%s`") at ink_error.cc:99 #3 0x77bb609e in _ink_assert (expression=0x800058 "(event == VC_EVENT_WRITE_COMPLETE) || (event == VC_EVENT_EOS)", file=0x7fe78c "HttpSM.cc", line=2763) at ink_assert.cc:37 #4 0x005f58cd in HttpSM::tunnel_handler_post (this=0x2aaabfcbe8f0, event=105, data=0x2aaabf009620) at HttpSM.cc:2763 #5 0x005f534b in HttpSM::main_handler (this=0x2aaabfcbe8f0, event=105, data=0x2aaabf009620) at HttpSM.cc:2665 #6 0x0051a5c8 in Continuation::handleEvent (this=0x2aaabfcbe8f0, event=105, data=0x2aaabf009620) at /usr/local/src/trafficserver/iocore/eventsystem/I_Continuation.h:153 #7 0x007a8235 in write_signal_and_update (event=105, vc=0x2aaabf009480) at UnixNetVConnection.cc:176 #8 0x007ac11a in UnixNetVConnection::mainEvent (this=0x2aaabf009480, event=1, e=0x2b5feb10) at UnixNetVConnection.cc:1289 #9 0x0051a5c8 in Continuation::handleEvent (this=0x2aaabf009480, event=1, data=0x2b5feb10) at /usr/local/src/trafficserver/iocore/eventsystem/I_Continuation.h:153 #10 0x007a10b6 in NetHandler::_close_vc (this=0x72409e50, vc=0x2aaabf009480, now=1486031338941225970, handle_event=@0x2b5fec28: 0, closed=@0x2b5fec2c: 0, total_idle_time=@0x2b5fec24: 0, total_idle_count=@0x2b5fec20: 0) at UnixNet.cc:679 #11 0x007a09a6 in NetHandler::manage_active_queue (this=0x72409e50, ignore_queue_size=true) at UnixNet.cc:588 #12 0x007a1cbb in InactivityCop::check_inactivity (this=0x73ecd100, event=2, e=0x2bc05a00) at UnixNet.cc:94 #13 0x0051a5c8 in Continuation::handleEvent (this=0x73ecd100, event=2, data=0x2bc05a00) at /usr/local/src/trafficserver/iocore/eventsystem/I_Continuation.h:153 #14 0x007cc28c in EThread::process_event (this=0x72406000, e=0x2bc05a00, calling_code=2) at UnixEThread.cc:143 #15 0x007cc6c7 in EThread::execute (this=0x72406000) at UnixEThread.cc:225 #16 0x007cb890 in spawn_thread_internal (a=0x73ec9040) at Thread.cc:84 #17 0x75a9ddc5 in start_thread (arg=0x2b5ff700) at pthread_create.c:308 #18 0x750af73d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:113 (gdb) frame 4 #4 0x005f58cd in HttpSM::tunnel_handler_post (this=0x2aaabfcbe8f0, event=105, data=0x2aaabf009620) at HttpSM.cc:2763 2763 ink_assert((event == VC_EVENT_WRITE_COMPLETE) || (event == VC_EVENT_EOS)); (gdb) print event $1 = 105 ``` --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1408: Prune some unused library dependencies
Github user zwoop commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/1408#discussion_r99174413 --- Diff: iocore/cache/Cache.cc --- @@ -1041,12 +1041,12 @@ CacheProcessor::cacheInitialized() case CACHE_COMPRESSION_FASTLZ: break; case CACHE_COMPRESSION_LIBZ: -#if !TS_HAS_LIBZ +#ifndef HAVE_ZLIB_H --- End diff -- So, while I'm generally OK with this (I think), I'm slightly concerned that we don't have a universal pattern for the #if vs #ifdef. In general, I think we ought to prefer #if over $ifdef's, assuming it's defined to "1". This helps avoid the problem where #define foo 0 is actually "defined", but is off. I know that we have plenty of cases where we do #ifndef / #ifdef, but unless there are other reasons to use that here, I think we should stick to #if when possible for these boolean types. Ideally, we'd later change this to be consistent as well. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1409: Log crc field and psql field correctly wit...
Github user zwoop closed the pull request at: https://github.com/apache/trafficserver/pull/1409 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1398: Configure option to disable CURL
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1398 Testing to see if clang-analyzer builds triggers [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1404: TS-1403 retry safe methods in case of serv...
Github user zwoop commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/1404#discussion_r98966326
--- Diff: proxy/http/HttpTransact.cc ---
@@ -6487,9 +6487,15 @@ HttpTransact::is_request_valid(State *s, HTTPHdr
*incoming_request)
bool
HttpTransact::is_request_retryable(State *s)
{
+ // If safe requests are retryable, it should be safe to retry safe
requests irrespective of bytes sent or connection state
+ // according to RFC the following methods are safe
(https://tools.ietf.org/html/rfc7231#section-4.2.1)
// If there was no error establishing the connection (and we sent
bytes)-- we cannot retry
- if (s->current.state != CONNECTION_ERROR &&
s->state_machine->server_request_hdr_bytes > 0 &&
- s->state_machine->get_server_session()->get_netvc()->outstanding()
!= s->state_machine->server_request_hdr_bytes) {
+ if (!(s->txn_conf->safe_requests_retryable == 1 &&
--- End diff --
Nit picky, but just treat is as a boolean? I.e. why explicitly compare it
with "1" ? We generally don't do that for other similar configs.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
[GitHub] trafficserver issue #1357: TS-5022: reduce memory allocation in clientcert l...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1357 @SolidWallOfCode Is this going to go in for 7.1.x ? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1403: idempotent requests should be retryable irrespect...
Github user zwoop closed the issue at: https://github.com/apache/trafficserver/issues/1403 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1400: traffic_via depends on the PCRE library
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1400 Testing again [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1400: traffic_via depends on the PCRE library
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1400 Testing again [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1403: idempotent requests should be retryable irrespect...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/issues/1403 Closing, since we have a PR already. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver pull request #1404: TS-1403 retry safe methods in case of serv...
Github user zwoop commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/1404#discussion_r99178025
--- Diff: proxy/http/HttpTransact.cc ---
@@ -6487,9 +6487,15 @@ HttpTransact::is_request_valid(State *s, HTTPHdr
*incoming_request)
bool
HttpTransact::is_request_retryable(State *s)
{
+ // If safe requests are retryable, it should be safe to retry safe
requests irrespective of bytes sent or connection state
+ // according to RFC the following methods are safe
(https://tools.ietf.org/html/rfc7231#section-4.2.1)
// If there was no error establishing the connection (and we sent
bytes)-- we cannot retry
--- End diff --
So, not 100% sure, but the original code did this:
```
// If the connection was established-- we cannot retry
if (s->state_machine->server_request_hdr_bytes > 0) {
```
Meaning, it would only retry if server_request_hdr_bytes is > 0. I'm not
sure what exactly that catches, but it seemed to imply that it'd only retry as
long as it wasn't established? But what does that mean, and what's the relation
to server_request_hdr_bytes?
The reason I'm asking now is that it's returnable regardless of these
"states", right? All it's checking is that the method is safe, and the
configuration allows retrying safe tests? What I'd like to see is that we get
back to the original behavior as much as possible, with the addition of
checking for safe methods before retrying.
Finally, I wonder if it'd make sense to create a isMethodSafe() function
somewhere? For example, I see similar checks in
```
// draft-stenberg-httpbis-tcp recommends only enabling TFO on safe,
indempotent methods or
// those with intervening protocol layers (eg. TLS).
if (scheme_to_use == URL_WKSIDX_HTTPS || t_state.method ==
HTTP_WKSIDX_CONNECT || t_state.method == HTTP_WKSIDX_DELETE ||
t_state.method == HTTP_WKSIDX_GET || t_state.method ==
HTTP_WKSIDX_HEAD || t_state.method == HTTP_WKSIDX_PUT) {
opt.f_tcp_fastopen = (t_state.txn_conf->sock_option_flag_out &
NetVCOptions::SOCK_OPT_TCP_FAST_OPEN);
}
```
which makes me think this should be a function? Particularly since this
looks wrong, PUT isn't safe. Also note that the comment above is wrong, if you
change this code, it should be "idempotent" here as well :). And afaik, all
safe methods are also idempotent, no?
I don't know if we check for safe methods elsewhere, but worth to take a
look. If I had a choice, I'd probably create two methods just to get this
properly documented in code:
```C
bool isMethodSafe(int method);
bool isMethodIdempotent(int method);
```
And link to the RFC 7231 as well here instead.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
[GitHub] trafficserver issue #1404: TS-1403 retry safe methods in case of server fail...
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1404 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1401: Segfault in write_to_net_io with 7.1.x
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/issues/1401 Are you getting a core file / backtrace ? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] trafficserver issue #1398: Configure option to disable CURL
Github user zwoop commented on the issue: https://github.com/apache/trafficserver/pull/1398 Testing to see if clang-analyzer builds triggers [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
